5-HT reuptake blockade induces pyroptosis in BRAF^V600E-mutated melanomas via remodeling histone serotonylation

Summary

The dual challenges of limited therapeutic options due to de novo or acquired resistance and psychological distress in patients with melanoma necessitate innovative treatment strategies. Here, we identify paroxetine hydrochloride (PH), a Food and Drug Administration (FDA)-approved antidepressant, as an alternative therapeutic for BRAF^V600E-mutated melanoma, including BRAFi/MEKi-resistant cases. Furthermore, our findings reveal that PH acts as an unrecognized inducer of pyroptosis. By triggering pyroptosis, PH remodels the tumor-permissive microenvironment in recurrent melanoma to potentiate anti-PD-1 therapy while maintaining a favorable safety profile. Mechanistically, PH impedes 5-hydroxytryptamine (5-HT) reuptake, leading to epigenetic reprogramming by reducing histone serotonylation (H3Q5ser) at the promoters of DNA repair genes. Impaired DNA damage repair pathways in turn trigger genome instability, proteostasis imbalance, and subsequent endoplasmic reticulum stress, ultimately inducing pyroptosis. Our findings uncover the underlying mechanism by which 5-HT drives melanoma progression and highlight PH as a promising candidate with multiple clinical potentials for treating melanoma.

Graphical abstract

Highlights

• •PH induces pyroptosis in melanoma harboring BRAF^V600E mutation
• •PH potentiates anti-PD-1 therapy in BRAFi/MEKi-resistant melanoma
• •5-HT is essential for melanoma cell survival by maintaining genomic stability

Li et al. demonstrate that paroxetine hydrochloride, an FDA-approved antidepressant, induces pyroptosis in melanoma and drives anti-tumor immunity. This occurs as 5-HT reuptake inhibition reduces H3Q5ser at DNA repair gene promoters, thereby impairing DNA damage repair, triggering genomic instability, disrupting proteostasis, and ultimately leading to pyroptosis.

Introduction

Melanoma, particularly when driven by BRAF^V600E mutations, represents a major challenge in the field of oncology because of its aggressive nature and resistance to conventional therapeutic modalities.¹ Despite the significant advancements in the survival and quality of life of patients with advanced disease resulting from the use of immune checkpoint inhibitors (ICIs), either alone or in combination, as well as targeted inhibitors of BRAF and MEK kinases in the mitogen-activated protein kinase pathway,² the occurrence of drug resistance remains a persistent challenge.^3,4,5 Therefore, the development of alternative therapeutic strategies is urgently needed.

Drug repurposing involves identifying the unrecognized applications of existing or experimental pharmaceuticals that deviate from their original medical indications.⁶ This approach offers multiple advantages over the development of a new drug, including a reduced risk of failure, a shortened drug development timeline, and a lower financial investment.⁷ In the context of cancer treatment, several drugs have been repurposed and successfully applied clinically, including thalidomide⁸ and losartan.^9,10,11 In addition, there has been a notable influx of drugs approved by the Food and Drug Administration (FDA) that have been validated in preclinical studies as promising options for cancer treatment,^{12,13,14,15,16,17,18,19} highlighting that drug repurposing is an important strategy for developing alternative therapeutic options for melanomas harboring the BRAF^V600E mutation.

Given that individuals with a cancer diagnosis frequently present with depressive symptoms as an accompanying condition, often as a consequence of the considerable distress they experience,²⁰ we were curious about whether a specific antidepressant could prove to be an effective intervention for both conditions. Here, we conducted a high-throughput drug screening based on 90 FDA-approved antidepressant drugs in melanoma cells and identified paroxetine hydrochloride (PH) as a promising therapeutic for melanoma harboring the BRAF^V600E mutation, with multiple clinical potentials. Furthermore, our findings revealed that PH treatment impeded the reuptake of 5-hydroxytryptamine (5-HT) (also known as serotonin), remodeled histone serotonylation, and consequently inhibited the signaling pathways associated with DNA damage repair (excision repair, homologous recombination repair, and mismatch repair). This ultimately resulted in genomic instability and the subsequent accumulation of unfolded proteins, which in turn led to endoplasmic reticulum (ER) stress and ultimately triggered pyroptosis. These findings highlight the potential of PH in the treatment of patients with BRAF^V600E-mutated melanoma.

Results

High-throughput drug screening reveals PH as an alternative therapeutic for BRAF^V600E-mutant melanoma

Patients with advanced melanoma are often clinically diagnosed with depression, prompting us to explore whether FDA-approved antidepressants exhibit anti-melanoma activity. To address this, we conducted high-throughput drug screening of 90 FDA-approved antidepressants in melanoma cells harboring the BRAF^V600E mutation. Among these candidates, four drugs, including PH, paroxetine hydrochloride hemihydrate, sertraline hydrochloride, and vortioxetine, exhibited strong anti-melanoma activity (inhibition rates >85%) (Figure 1A). In addition to PH showing the greatest effectiveness in inhibiting the viability of melanoma cells during our screening process, pretreatment with PH has been demonstrated to alleviate depressive symptoms in patients with malignant melanoma.²¹ Therefore, we focused on the potential role of PH in suppressing melanoma proliferation. We investigated the potential anti-melanoma effects of PH by assessing its IC₅₀ in multiple BRAF^V600E-mutated melanoma cell lines (Figure 1B) and confirmed its dose-dependent antiproliferative effects in these melanoma cell lines (Figures 1C and S1A). More importantly, after establishing the appropriate in vivo concentration of PH (Figures S1B–S1E), we further validated its tumor-suppressive activity and favorable safety profile in melanoma xenograft models (Figures 1D–1F and S1F). In addition, PH treatment only marginally affected the viability of nine types of normal cells derived from different organs, again highlighting its favorable safety profile (Figure S1G).

Figure 1.

High-throughput drug screening reveals PH as an alternative therapeutic for BRAF^V600E-mutant melanoma

(A) Waterfall plots showing the effects of an antidepressant compound library containing 90 drugs. Inhibition rates >85% drugs (green). Three biological replicates were screened for each sample.

(B) IC₅₀ of PH in the indicated cell lines.

(C) Colony formation assays in the indicated cells treated with DMSO or gradient doses of PH.

(D) A375 xenografts were established and treated with vehicle or PH (25 mg/kg daily intraperitoneally [i.p.]) for 14 days. Tumors in each group were individually recorded every 2 days. n = 8 tumors per group.

(E and F) Weight (E) and image (F) of xenograft tumors in (D).

(G) 5-HT levels in A375 cells treated with DMSO or the indicated PH treatment for 24 h.

(H) 5-HT levels in A375 cells with the indicated treatment.

(I and J) CCK8 (I) and clonogenic (J) assays in A375 cells with the indicated treatment.

Data in (D) and (E) are displayed as mean ± SEM, and the others are displayed as mean ± SD. Statistical significance was determined using Student’s t test. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05.

See also Figure S1.

Considering the established role of PH in the suppression of 5-HT reuptake and GRK activation, we next examined the potential mechanisms of action of PH. Despite the reduction in GRK expression observed with increasing doses of PH (Figure S1H), GRK silencing did not alter the proliferation rate of melanoma cells (Figures S1I and S1J). This led to the hypothesis that the anti-melanoma function of PH was primarily dependent on the inhibition of 5-HT reuptake. Additionally, it was supported by the observation that PH reduced the intracellular 5-HT levels in a dose-dependent manner (Figure 1G). To further confirm 5-HT dependency in PH-induced proliferation defects, we applied increasing doses of exogenous 5-HT to PH-treated cells (Figure 1H). The restoration of intracellular 5-HT significantly enhanced melanoma cell proliferation after PH treatment (Figures 1I and 1J). However, these findings raised an interesting question: if melanoma cells were capable of synthesizing 5-HT on their own, why did they still rely on the reuptake of extracellular 5-HT to sustain their viability? Intriguingly, after analyzing the expression distribution of TPH1 (encoding the rate-limiting enzyme for 5-HT synthesis) in A375 cells, we found that only a small portion of melanoma cells highly expressed TPH1 (Figure S1K), suggesting that not all of them synthesize 5-HT. To further explore whether a subpopulation may act as 5-HT producers, supporting the surrounding cell population, we generated A375 cells that stably overexpressed GFP-TPH1 (namely, GFP-A375) or mCherry (mCherry-A375) and thereafter performed the experiment as shown in Figure S1L. Of note, under PH treatment, intracellular 5-HT levels were significantly higher in mCherry-labeled cells grown in co-culture than in those cultured alone (Figure S1M). Consistently, the co-cultured cells were more resistant to PH treatment (Figure S1N). These findings suggest a metabolic cooperation mechanism within melanoma cells, in which a subpopulation serves as a 5-HT producer to increase intracellular 5-HT levels in surrounding cells and thereby support their survival. Taken together, these findings reveal that PH reduces the 5-HT reuptake to attenuate melanoma progression and indicate a specific role of 5-HT in maintaining melanoma proliferation.

PH suppresses BRAFi/MEKi-resistant melanoma progression

Patients with recurrent tumors often experience heightened psychological stress and are more prone to depression. Therefore, we investigated the potential of PH for managing recurrent melanoma. We established multiple BRAFi (dabrafenib) and MEKi (trametinib)-resistant (hereafter named DTR) melanoma cells by treating them with BRAFi/MEKi (DT) for approximately 3 months (Figures S2A–S2C). Intriguingly, slightly increased sensitivity to PH was observed in multiple DTR melanoma cells compared to the parent cell line, as evidenced by the reduced IC₅₀ values (Figure 2A). Meanwhile, PH also suppressed the proliferation of DTR melanoma cells in a dose-dependent manner (Figures 2B and S2D). Given that minimizing pharmacologic burden generally reduces treatment-related adverse events, we sought to verify whether PH monotherapy could serve as a potential therapeutic option for patients with recurrent melanoma using a xenograft model. Notably, DT treatment was independent of maintaining the drug resistance phenotype in vivo, as evidenced by the findings that xenografts derived from A375^DTR cells exhibited similar proliferation rates regardless of DT treatment status (Figures S2E–S2H), providing an opportunity to explore the therapeutic potential of PH monotherapy in recurrent melanoma. Prior to initiating the in vivo investigation, we established the optimal dose of PH (Figures S2I–S2L). Similar to its effect on parental tumors, PH at 25 mg/kg markedly suppressed the growth of DTR tumors while exhibiting a favorable safety profile (Figures 2C–2E and S2M).

Figure 2.

PH inhibits BRAFi/MEKi-resistant melanoma progression

(A) IC₅₀ of PH in the indicated DTR cell lines.

(B) Colony formation assays in the indicated DTR cells treated with DMSO or the indicated doses of PH.

(C) A375^DTR xenografts were established and treated with vehicle or PH (25 mg/kg daily i.p. injection) for 15 days. The tumors in each group were individually recorded every 3 days. n = 8 tumors per group.

(D and E) Weight (D) and image (E) of the xenograft tumors in (C).

(F) 5-HT levels inA375^DTR cells treated with DMSO or indicated doses of PH for 24 h.

Data in (C) and (D) are displayed as mean ± SEM, and (F) is displayed as mean ± SD. Statistical significance was determined by Student’s t test. ∗∗∗p < 0.001, ∗∗p < 0.01.

See also Figure S2.

Consistent with the observations in parental cells, GRK expression was reduced by PH treatment in a dose-dependent manner in DTR cells (Figure S2N). Similarly, GRK depletion also failed to suppress DTR melanoma proliferation, whereas PH treatment drastically reduced intracellular 5-HT levels (Figures 2F, S2O, and S2P), indicating that PH shares a mechanism for its antiproliferative effects in parental and BRAFi/MEKi-resistant melanoma. Based on this, we proceeded to investigate the potential explanation for why DTR cells are slightly more sensitive to PH treatment. Firstly, we compared the expression levels of TPH1 and SLC6A4 (encoding the 5-HT transporter) between parental and DTR cells. Unexpectedly, TPH1 expression was substantially lower in DTR cells compared to parental cells, while SLC6A4 was slightly downregulated (Figure S2Q). Consistent with the reduced TPH1 levels, intracellular 5-HT was also significantly decreased in DTR cells (Figure S2R). Moreover, intracellular 5-HT levels were found to be lower in DTR cells compared to parental cells following PH treatment (Figure S2R). Hence, we hypothesized that in DTR cells, due to the decreased expression of TPH1, intracellular 5-HT levels were at the critical threshold required for maintaining normal cell growth. As a result, once the cells underwent PH treatment, intracellular 5-HT levels would drop to a lower level, thereby making DTR cells more sensitive than parental cells. Additionally, due to the TPH1 reduction, DTR cells might rely on hyperactivated SLC6A4 to maintain the critical threshold level of intracellular 5-HT required for cell growth. To verify this hypothesis, we silenced TPH1 in parental cells, adjusting its expression to a level comparable to that in DTR cells (Figure S2S). Surprisingly, when TPH1 expression reached comparable levels, intracellular 5-HT levels in DTR cells were significantly higher than in parental cells (Figure S2T), indicating that DTR cells depended on enhanced SLC6A4 activity to sustain 5-HT homeostasis, rendering them more sensitive to PH. Collectively, these findings again highlight the dual clinical potential of PH in melanoma management.

PH inhibits melanoma progression via triggering pyroptosis

While exploring how PH exerts its antiproliferative effects, we found that PH treatment led to significant cell death with morphological features consistent with pyroptosis in both parental and DTR cells, as they exhibited swelling and extensive plasma membrane blebbing (Figures 3A and S3A). This observation was accompanied by membrane rupture, as indicated by drastically elevated lactate dehydrogenase (LDH) levels after PH treatment (Figures 3B and S3B), and a dramatic increase in the propidium iodide/Annexin V positive (PI⁺/Annv⁺) cell population (Figures 3C, S3C, and S3D). These data suggest that PH induces lytic cell death accompanied by a pyroptotic morphology in both parental and DTR melanoma cells.

Figure 3.

PH acts as an unrecognized pyroptosis inducer

(A) A375 cells were treated with PH (15 μM) for 24 h to assess the characteristic morphology. The red arrows indicate pyroptotic cells. Scale bars, 100 μm.

(B and C) LDH release (B) and percentage of Annexin V (Annv)⁺/propidium iodide (PI)⁺ cell population (C) at different concentrations of PH for 24 h in A375 cells.

(D) GSDMB cleavage at different concentrations of PH for 12 h in A375 cells. GSDMB-FL, GSDMB full length; GSDMB-N, N terminus of the cleaved GSDMB.

(E) GSDMB cleavage in the indicated parental melanoma cells with different concentrations of PH treatment for 12 h.

(F) A375 cells were treated with DMSO or PH (10 μM) combined with the apoptosis inhibitor Z-DEVD-FMK (40 μM), necroptosis inhibitor necrosulfonamide (2 μM), or ferroptosis inhibitor ferrostatin-1 (10 μM) for 48 h and then subjected to CCK-8 assay.

(G) GSDMB cleavage in A375 cells stably expressing the indicated short hairpin RNA (shRNA) following the indicated treatment.

(H and I) CCK-8 (H) and clonogenic (I) assays in A375 cells stably expressing indicated shRNA following the indicated treatment.

(J) A375 cells stably expressing the indicated shRNA were treated with PH for 24 h to assess the characteristic morphology. The red arrows indicate pyroptotic cells. Scale bars, 100 μm.

(K and L) LDH release (K) and percentage of Annv⁺/PI⁺ cell population (L) in A375 cells stably expressing the indicated lentiviral constructs following PH treatment.

(M–O) A375 cells expressing indicated lentiviral constructs were subcutaneously injected into BALB/c nude mice. When tumor sizes reached 50 mm³, mice were intraperitoneally (i.p.) injected with vehicle or PH (25 mg/kg) once daily. The tumor size in each group was recorded every 2 days. At the endpoint, tumors were excised and then photographed (N) and weighed (O). n = 8 tumors per group.

Data in (B), (C), (F), (H), (I), (K), and (L) are displayed as mean ± SD. Other data are displayed as mean ± SEM. Statistical significance was determined by Student’s t test. ∗∗∗p < 0.001, ∗∗p < 0.01, n.s., not significant.

See also Figures S3 and S4.

We further sought to confirm whether PH is an inducer of pyroptosis. Cleaved gasdermin family proteins are the executors of pyroptosis.²² Among gasdermin family members, PH treatment induced the cleavage of GSDMB in a dose-dependent manner in multiple parental melanoma cells (Figures 3D and 3E), while no cleavage of GSDMA, GSDMC, GSDMD, or GSDME was observed (Figures S3E–S3H). In addition, dose-dependent of PH-induced GSDMB cleavage was also observed in the corresponding DTR cells (Figures S3I–S3K). Meanwhile, PH-induced pyroptosis-like cell death was not reversed by pre-treatment with Z-DEVD-FMK (an inhibitor of apoptosis through the inhibition of caspase-3), necrosulfonamide (an inhibitor of necroptosis through the inhibition of MLKL), and ferrostatin-1 (a ferroptosis inhibitor), further indicating PH-induced pyroptosis (Figure 3F). To gain a mechanistic insight, we constructed GSDMA-, GSDMB-, GSDMC-, GSDMD-, and GSDME-depleted A375 cells and treated these cells with PH. Notably, depletion of GSDMB, but not other GSDM members, resumed cell proliferation (Figures 3G–3I and S3L–S3N). Furthermore, the pyroptotic morphological alterations induced by PH treatment were largely abolished by GSDMB silencing, whereas depletion of other GSDM members failed to do so (Figures 3J and S4A), indicating that PH induces GSDMB-dependent pyroptosis. To further explore our hypothesis, we examined whether GSDMB inhibition could rescue the PH-induced LDH release and increase of Annv⁺/PI⁺ rates. Unsurprisingly, the increased LDH release and Annv⁺/PI⁺ rates following PH treatment were also drastically reversed upon depletion of GSDMB (Figures 3K, 3L, and S4B). Similar to these in vitro findings, GSDMB depletion resumed tumor growth in both A375 and A375^DTR xenograft models (Figures 3M–3O and S4C–S4G), demonstrating GSDMB dependency in vivo. Collectively, these results reveal that PH induces GSDMB-dependent pyroptosis to exert its anti-melanoma functions.

To further confirm whether PH-induced pyroptosis relies on the inhibition of 5-HT reuptake, we restored intracellular 5-HT levels (Figure 1H). Notably, the pyroptotic morphological alterations induced by the PH treatment were diminished following the restoration of 5-HT levels (Figure S4H). Consistently, supplementation with exogenous 5-HT also remarkably reduced LDH release and the proportion of PI⁺/Annv⁺ cells following PH treatment (Figures S4I and S4J). More importantly, cleaved GSDMB induced by PH treatment was hardly observed after the restoration of intracellular 5-HT levels (Figure S4K). Taken together, these findings demonstrate that PH suppresses 5-HT reuptake to induce GSDMB-dependent pyroptosis, thereby exerting its anti-melanoma functions.

PH potentiates anti-PD-1 immunotherapy in DTR melanoma

Acquired resistance to BRAFi/MEKi generates an immune-evasive tumor microenvironment (TME) and confers cross-resistance to immunotherapy in melanoma.^23,24 Considering that pyroptosis is recognized as a type of immunogenic cell death,²⁵ we therefore set out to explore another potential clinical application of PH in melanoma treatment: whether it potentiates PD-1 blockade in recurrent melanomas characterized by low immunogenicity.

As established in our previous study, an immunosuppressive TME was observed in a syngeneic mouse model bearing YUMM1.7^DTR cells,²⁶ evidenced by the significantly decreased staining intensity of both CD8 and granzyme B (GZMB) (Figure 4A). To further explore our hypothesis, we first examined whether PH induced pyroptosis in DTR murine melanoma cells. Consistent with the results obtained in human melanoma cells, PH treatment reduced intracellular 5-HT levels in YUMM1.7^DTR cells in a dose-dependent manner (Figure S5A). Similarly, the growth rate of YUMM1.7^DTR cells was inhibited following PH treatment (Figures S5B and S5C). Furthermore, we confirmed that PH treatment efficiently induced pyroptosis, as evidenced by the emergence of typical pyroptotic cell death morphology, elevated LDH release, and increased proportion of Annv⁺/PI⁺ cells (Figures S5D–S5F). Of note, murine cells reportedly lack the expression of GSDMB.^27,28 We speculated that PH treatment triggers pyroptosis by activating other GSDM molecules. As shown in Figures S5G and S5H, cleavage of GSDMC, but not GSDMA, GSDMD, or GSDME, was observed in YUMM1.7^DTR cells following PH treatment, indicating that PH might induce GSDMC activation to trigger pyroptosis in these cells. To confirm GSDMC dependency, we generated GSDMA-, GSDMC-, GSDMD-, and GSDME-depleted YUMM1.7^DTR cells (Figure S5I). Depletion of GSDMC, but not other GSDM members, restored cell proliferation (Figures S5J and S5K). Meanwhile, pyroptotic morphological alterations were observed in GSDMA-, GSDMD-, and GSDME-depleted cells but not in GSDMC-silenced cells following PH treatment (Figure S5L). Moreover, once GSDMC cleavage was blocked by its depletion (Figure S6A), the increased LDH release and PI⁺/Annv⁺ rates induced by PH treatment were drastically reversed (Figures S6B and S6C). Collectively, these findings demonstrate that PH induces GSDMC-dependent pyroptosis in YUMM1.7^DTR cells.

Figure 4.

PH induces pyroptosis to potentiate PD-1 blockade

(A) Representative immunofluorescence images of CD8 and GZMB staining in YUMM1.7^P- and YUMM1.7^DTR-derived tumors; right: quantification of indicated staining-positive cells. Scale bars, 10 μm.

(B–D) YUMM1.7^DTR cells stably expressing the indicated shRNAs were subcutaneously injected into immunocompetent C57BL/6J mice. When tumor sizes reached 50 mm³ (day 5, labeled as day 0 in C), mice were administrated with the indicated drugs (B). Tumor size in each group was recorded every 3 days (C). At the endpoint, the tumors were excised and weighed (D). PH, 25 mg/kg, intraperitoneal (i.p.) injection, daily. IgG2a or anti-PD-1 was diluted in saline and then given i.p. every 3 days. n = 7 tumors per group.

(E) CD8 and GZMB staining in YUMM1.7^DTR allografts with the indicated treatment (left). Right: quantification of the indicated staining-positive cells. Scale bars, 10 μm.

(F) Frequencies of intratumoral CD3ε⁺ (left), CD4⁺ (middle), and CD8⁺ (right) cells among CD45⁺ cells. n = 7 tumors per group.

(G) Frequencies of intratumoral GZMB⁺ (left), PRF1⁺ (middle), and IFN-γ⁺ (right) cells among CD8⁺ T cells. n = 7 tumors per group. Data are displayed as mean ± SEM. Statistical significance was determined by Student’s t test. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, n.s., not significant.

See also Figures S5, S6, and S7.

Next, we explored whether PH treatment could potentiate PD-1 blockade by boosting immunogenic pyroptosis (Figure 4B). In YUMM1.7^DTR allografts, PH monotherapy significantly impaired tumor growth in vivo, whereas anti-PD-1 monotherapy failed to do so (Figures 4C, 4D, and S6D). Of note, the addition of anti-PD-1 to PH produced a further significant, albeit not striking as expected, reduction in tumor burden. Notably, once the cleavage of GSDMC was blocked by its depletion (Figure S6E), tumors became insensitive to both PH monotherapy and the combination therapy, underscoring GSDMC dependency. More importantly, there were no signs of off-tumor toxicity as evidenced by the absence of histological lesions in the major organs, including the heart, lungs, kidneys, and liver, or no weight loss (Figures S6F and S6G). Collectively, these data suggest that PH-induced pyroptosis is the primary driver of tumor control in this model, with the combination conferring an incremental benefit.

To further investigate whether PH treatment instructs the TME by activating GSDMC-dependent pyroptosis, we conducted immunofluorescence staining, and a marked influx of CD8⁺ T cells was observed following PH administration in YUMM1.7^DTR tumors (Figure 4E). Flow cytometry confirmed the increased tumor-infiltrating CD3ε⁺, CD8⁺, and CD4⁺ T cells in tumors exposed to PH treatment (Figures 4F and S7A). These findings suggest that blocking 5-HT reuptake drives anti-tumor immunity in recurrent melanomas. In contrast, upon GSDMC ablation, no significant alterations in infiltrated CD8⁺ T cells were observed in tumors subjected to PH monotherapy or PH/anti-PD-1 co-administration (Figures 4E and 4F), demonstrating that GSDMC-mediated pyroptosis is indispensable for PH-mediated TME remodeling in DTR melanoma tumors. Moreover, PH/anti-PD-1 co-administration stimulated tumor-infiltrating CD8⁺ T cell activation, as evidenced by the increased GZMB staining (Figure 4E). In line with this notion, flow cytometry analysis of tumor tissues revealed significantly increased interferon (IFN)-γ, perforin 1 (PRF1), and GZMB-positive rates in CD8⁺ T cells purified from PH/anti-PD-1 co-administered tumors in a GSDMC-dependent manner, suggesting that combination therapy exerts its anti-tumor activity by strengthening the tumor-killing effector function of CD8⁺ T cells (Figures 4G and S7A). Notably, no significant changes in the proportions of CD3ε⁺, CD8⁺, and CD4⁺ T cells were found in the spleens and peripheral blood of mice following PH/anti-PD-1 co-administration, suggesting that PH or combination therapy specifically changes the TME without affecting the peripheral immunity (Figures S7B–S7E). Collectively, these findings indicate an additional promising clinical application of PH that potentiates anti-PD-1 therapy in recurrent BRAF^V600E-mutated melanomas.

PH disturbs the protein homeostasis to induce ER stress

To gain a mechanistic insight into PH-triggered pyroptosis, we profiled the transcriptional signature after the PH treatment. As shown in Figure S8A, a total of 1,668 downregulated genes and 1,400 upregulated genes were identified upon PH treatment. Pathway analysis revealed that the upregulated genes were majorly enriched in ER stress-related biological processes (Figure S8B). Given that ER stress is considered a key driver of pyroptosis,²⁹ we therefore conducted a more detailed analysis of the transcriptional alterations associated with various elements of the ER stress signal transduction pathway using gene set enrichment analysis (GSEA). Of note, we found that ER stress signaling was transcriptionally activated at every step from upstream to downstream, suggesting that PH treatment may disrupt the proteostasis to induce ER stress (Figure 5A). Therefore, we monitored protein aggregation following PH treatment using PROTEOSTAT, a molecular rotor dye that specifically interacts with the cross-β spines of the quaternary protein structures present in misfolded and aggregated proteins. As expected, PH treatment induced protein aggregation in both parental and BRAFi/MEKi-resistant cells (Figures 5B and S8C). Meanwhile, we found that PH treatment in parental and DTR cells led to an excessive activation of ATF6 but not PERK and IRF in these cells, as supported by the increased cleavage and nuclear translocation of ATF6 (Figures 5C, 5D, S8D, and S8E). Given that the reduced protein synthesis typically occurs during ER stress,³⁰ we used a puromycin incorporation assay to measure protein synthesis activity. As expected, the protein synthesis was inhibited in both parental and DTR melanoma cells following PH treatment, as evidenced by the decreased puromycin intensity (Figures 5E and S8F). Collectively, these findings reveal that PH treatment disrupts proteostasis.

Figure 5.

PH disrupts proteostasis to trigger ER stress

(A) GSEA of ER stress-related pathways in PH-induced genes.

(B) Representative images (left) and quantification (right) of PROTEOSTAT (magenta) and DAPI (blue) staining treated with DMSO or PH for 24 h in A375 cells. Scale bars, 15 μm.

(C) IB analysis of whole-cell lysates derived from A375 cells treated with different concentrations of PH for 24 h (upper) and its quantification (lower).

(D) Representative images (left) and quantification (right) of ATF6 (red) and DAPI (blue) staining in A375 cells with different concentrations of PH treatment. Scale bars, 5 μm.

(E) A375 cells treated with different concentrations of PH were collected for IB analysis as indicated.

(F and G) A375 cells were treated with DMSO or PH combined with 4-PBA (1 μM) for 24 h, followed by IB analysis as indicated (left). Right: quantifications corresponding to the intensity of the indicated proteins.

(H–J) Effects of 4-PBA (1 μM) on PH-induced pyroptosis for characteristic morphology (H), GSDMB cleavage (I), and cell viability (J). Scale bars, 100 μm. Data are displayed as mean ± SD. Statistical significance was determined by Student’s t test. ∗∗∗p < 0.001, ∗∗p < 0.01, ∗p < 0.05, n.s., not significant.

See also Figures S8, S9, and S10.

We next explored whether PH-induced pyroptosis depended on triggering ER stress. 4-PBA, an ER stress inhibitor, recovered the excessive activation of ATF6 and reduced total protein synthesis induced by PH treatment in human melanoma cells (Figures 5F and 5G), confirming that 4-PBA blunted PH-induced ER stress. Of note, once PH failed to induce ER stress, the pyroptotic morphological alterations were largely diminished (Figure 5H). In line with this notion, 4-PBA treatment drastically reduced LDH release and PI⁺/Annv⁺ rates, along with diminished cleavage of GSDMB in PH-treated human melanoma cells (Figures 5I, S8G, and S9A). Consequently, 4-PBA treatment resumed cell proliferation after PH treatment (Figures 5J and S9B). Collectively, these data establish that PH treatment disrupts proteostasis to induce ER stress, thereby triggering pyroptosis.

Inspired by the fact that ER stress drives caspase-8 activation³¹ and activated caspase-8 cleaves GSDMC to trigger pyroptosis,^32,33 we confirmed that 4-PBA treatment was sufficient to abolish PH-induced pyroptotic morphological alterations, elevated LDH release, increased PI⁺/Annv⁺ rates, cleavage of GSDMC, and subsequent growth defects (Figures S9C–S9J), demonstrating that PH triggers ER stress to induce pyroptosis in murine melanoma cells. Based on these findings, we further explored whether caspase-8 activation is indispensable for PH-induced GSDMC cleavage and subsequent pyroptosis in YUMM1.7^DTR cells. Interestingly, PH induced caspase-8 activation in a dose-dependent manner (Figure S10A). Furthermore, when the activation of caspase-8 was blocked by its depletion (Figure S10B), the typical pyroptotic morphological alterations were largely diminished (Figure S10C). Consistently, caspase-8 silencing also remarkably reduced LDH release and PI⁺/Annv⁺ rates in YUMM1.7^DTR cells following the PH administration (Figures S10D and S10E). Consequently, caspase-8 depletion profoundly enhanced YUMM1.7^DTR cell proliferation under PH treatment (Figures S10F and S10G). Taken together, our findings demonstrate that, in murine melanoma cells, 5-HT reuptake blockage in murine melanoma cells triggers ER stress to activate caspase-8, thereby driving GSDMC-dependent pyroptosis.

PH impedes DNA damage repair signaling to drive genome instability

Given that 5-HT facilitates gene transcription by modifying H3 (H3Q5ser) and thereby stabilizing the enrichment of H3K4me3,³⁴ we focused on the downregulated genes, aiming to elucidate the direct mechanism by which 5-HT reuptake inhibition disrupts proteostasis. Notably, GSEA revealed that the inhibition of 5-HT reuptake blunted the DNA damage repair pathways, including base excision repair, homologous recombination repair, and mismatch repair (Figure 6A). In detail, PH treatment impeded the transcription of most genes in these pathways (Figure S11A). Next, we performed H3Q5ser chromatin immunoprecipitation (ChIP) sequencing (ChIP-seq) to explore how intracellular 5-HT regulates the transcription of these genes. As shown in Figure S11B, approximately 8.8% of the H3Q5ser peaks were localized at promoters, whereas 85.1% were detected at enhancer regions. Among these, PH treatment drastically reduced the enrichment of H3Q5ser in the promoters of the downregulated genes (Figure 6B). Given the role of H3Q5ser in stabilizing H3K4me3 and potentiating its readout,³⁴ we next performed H3K4me3-ChIP-seq. Although the genomic distribution of H3K4me3 differed from that of H3Q5ser, approximately 26.5% of H3K4me3 peaks were localized at promoters, whereas 63% were detected at enhancers (Figure S11C), and the enrichment among those downregulated genes was reduced following PH treatment (Figure S11D). Notably, a significant loss of H3Q5ser enrichment and accumulation of H3K4me3 marks were observed at promoters of key DNA repair-related genes, including POLD1, XRCC2, and FEN1 (Figure 6C). To validate our RNA sequencing (RNA-seq) and ChIP-seq data, we first performed ChIP-quantitative reverse-transcription PCR (RT-qPCR) and found a drastically decreased genomic deposition of H3Q5ser and H3K4me3 at the promoter regions of DNA repair-related genes, including POLD1, XRCC2, and FEN1, in both parental and DTR cells (Figures S11E–S11H). Consistently, immunoblotting (IB) and RT-qPCR revealed that the expression levels of these genes were remarkably decreased in PH-treated cells (Figures S11I–S11L). More importantly, the restoration of intracellular 5-HT levels significantly enhanced the genomic deposition of H3Q5ser and H3K4me3 at the promoter regions of these genes in PH-treated cells (Figures S11M and S11N). As a result, upon the restoration of 5-HT, the expression levels of these genes were also rescued (Figures S12A and S12B). Collectively, these findings indicate that 5-HT is a critical factor in maintaining genome stability by sustaining the transcription of DNA repair-associated pathways in melanoma cells.

Figure 6.

PH transcriptionally inactivates DNA damage repair to induce misfolded protein accumulation and subsequent pyroptosis

(A) GSEA of DNA damage repair-associated pathways, including base excision repair, homologous repair, and mismatch repair with PH-regulated genes.

(B) Normalized H3Q5ser tag densities at the promoter of the direct downstream genes (n = 534) in A375^P cells under DMSO or PH treatment.

(C) H3Q5ser and H3K4me3 ChIP-seq tracks surrounding the promoters of the indicated genes in A375^P cells.

(D) Representative images (left) and quantification (right) of γH2AX (green) and DAPI (blue) staining in A375^P cells treated with DMSO or PH for 24 h. Scale bars, 15 μm.

(E) A375^P cells stably expressing the indicated lentiviral constructs were treated with PH, followed by IB analysis as indicated.

(F) Representative images of γH2AX (green) and DAPI (blue) staining in A375^P cells stably expressing the indicated lentiviral constructs following PH treatment. Scale bars, 5 μm.

(G) Representative images (left) and quantification (right) of PROTEOSTAT (magenta) and DAPI (blue) staining in A375^P cells stably expressing the indicated lentiviral constructs following PH treatment. Scale bars, 15 μm.

(H) A375^P cells stably expressing indicated shRNA were treated with PH (15 μM) for 24 h to assess the characteristic morphology. The red arrows indicate pyroptotic cells. Scale bars, 100 μm.

(I) LDH release in A375^P cells stably expressing the indicated lentiviral constructs following PH treatment.

(J) A375^P cells stably expressing the indicated lentiviral constructs were treated with PH, followed by IB analysis as indicated.

(K) CCK-8 assay in A375^P cells stably expressing the indicated lentiviral constructs following PH treatment for 48 h. Data are displayed as mean ± SD. Statistical significance was determined by Student’s t test. ∗∗∗p < 0.001.

See also Figures S11, S12, S13, and S14.

To determine whether PH treatment indeed led to DNA damage accumulation in both parental and BRAFi/MEKi-resistant cells, we first conducted immunofluorescent staining and observed a drastically increased γ-H2AX following PH treatment (Figures 6D and S12C). This observation was further validated by IB analysis (Figures S12D and S12E). Furthermore, comet assay as well as 53BP1 staining further confirmed the occurrence of DNA damage in both parental and DTR cells following PH treatment (Figures S12F–S12I).

To investigate whether DNA damage is the prerequisite for PH-induced disruption of proteostasis and subsequent pyroptosis, we restored these three DNA repair genes, including XRCC2, POLD1, and FEN1 (Figure 6E). Restoration of these genes significantly reduced DNA damage accumulation in cells following PH treatment, as evidenced by the reduced γ-H2AX levels (Figures 6E and 6F). Notably, under this condition, the accumulation of misfolded proteins was drastically reduced when DNA damage was decreased (Figure 6G). As a consequence, the ER stress was alleviated as supported by the decreased cleavage and nuclear translocation of ATF6 (Figures S12J and S12K). Together, these findings demonstrate that PH treatment induces DNA damage to disrupt proteostasis, thereby triggering ER stress. Furthermore, the restoration of DNA repair-related genes also conferred resistance to PH-induced pyroptosis in melanoma cells, as evidenced by reduced pyroptotic morphological alterations, cleavage of GSDMB, LDH release, and the proportion of PI⁺/Annv⁺ cells (Figures 6H–6J and S13A). In line with this notion, cell proliferation was also significantly resumed following PH treatment when the expression of these genes was restored (Figures 6K and S13B). Meanwhile, similar findings were also observed in DTR cells (Figures S13C–S13H and S14A–S14E). Collectively, these data demonstrate that PH treatment blocks 5-HT reuptake, impeding DNA damage repair-associated pathways and thereby disrupting proteostasis, which ultimately leads to pyroptosis in melanoma cells.

Clinical implication of TPH1/SLC6A4 in BRAF^V600E-mutated melanomas

To explore the clinical implications of TPH1 and SLC6A4, two genes directly involved in maintaining the intracellular 5-HT pool, we conducted an analysis of melanoma patients harboring the BRAF^V600E mutation (n = 48) using The Cancer Genome Atlas (TCGA) database and then performed gene expression correlation analysis, and the genes that were positively correlated with TPH1 and SLC6A4 were independently identified (R > 0.3, p < 0.05; Figure 7A). Intriguingly, most genes that significantly correlated with SLC6A4 were positively associated, suggesting that SLC6A4 mediates 5-HT reuptake to facilitate gene transcription in BRAF^V600E-mutated melanomas. By overlapping these gene sets, we identified 1,718 genes that were significantly co-expressed with both TPH1 and SLC6A4. Further pathway enrichment analysis of these co-expressed genes highlighted DNA repair pathways as among the most significantly enriched categories (Figures 7B and 7C), thereby providing further support for our mechanistic findings.

Figure 7.

Clinical implications of TPH1/SLC6A4 expression in BRAF^V600E-mutated melanomas

(A) Scatterplot representing the differentially correlated genes with TPH1 (left) and SLC6A4 (right). Genes that were significantly correlated are highlighted in red (p < 0.05). p values were calculated by Fisher’s exact test.

(B and C) Gene Ontology analysis of genes exhibiting significantly positive correlation between TPH1 and SLC6A4.

(D) Uniform manifold approximation and projection (UMAP) representing all sequenced cells based on patient ID (left), the expression distribution of TPH1/SLC6A4 (middle), and the expression distribution of the pyroptosis signature (right) in these two samples. Quantification of pyroptosis signature expression in malignant cells from these two samples. p value was determined by Student’s t test.

(E) UMAP representing all sequenced cells based on cell type in patient #1 (left) and patient #2 (right).

(F) Heatmap showing the quantification of the number of the indicated cell types in patient #1 and patient #2 based on the expression of cell type specific genes.

(G–I) Scatterplot derived from the TCGA BRAF^V600E-mutated melanoma dataset (n = 48) illustrating the correlation between TPH1/SLC6A4 levels and the stromal score (G), immune score (H), and tumor purity (I). p values were calculated by Fisher’s exact test.

Importantly, using single-cell RNA-seq conducted on patients with BRAF^V600E-mutated melanoma, who are resistant to targeted therapy, revealed a clinically significant correlation with our conclusion. Specifically, one resistant patient exhibited a tumor with high TPH1/SLC6A4 expression, whereas the other patient had a tumor with low TPH1/SLC6A4 expression (Figure 7D). Interestingly, the transcription levels of the pyroptosis signature were remarkably lower in the tumor with high TPH1/SLC6A4 expression (Figure 7D), suggesting that melanomas with high 5-HT levels may be resistant to pyroptosis. Meanwhile, the tumor characterized by high TPH1/SLC6A4 expression demonstrated markedly reduced infiltration of CD8⁺ T cells and natural killer cells (Figures 7E and 7F), which is consistent with our conclusion that the blockade of 5-HT signaling drives anti-tumor immunity. Consistently, analyses of the TCGA BRAF^V600E melanoma cohort revealed a significant negative correlation between TPH1/SLC6A4 expression and both stromal score and immune score (Figures 7G and 7H), as well as a positive correlation with tumor purity (Figure 7I), underscoring that BRAF^V600E-mutated melanomas with highly activated 5-HT signaling are characterized by a less immunogenic TME, which may indicate reduced sensitivity to immune checkpoint blockade therapy. Collectively, these findings suggest that enhanced intracellular 5-HT levels may contribute to the formation of a tumor-permissive TME in BRAF^V600E melanomas.

Discussion

Patients with cancer, especially those with advanced or recurrent disease, frequently experience severe emotional distress that often manifests as clinical depression.²⁰ This presents a distinctive opportunity to explore therapies that address both psychological and oncological needs. Accordingly, we conducted a high-throughput screening of FDA-approved antidepressants in melanoma cells, resulting in the identification of PH, a selective 5-HT reuptake inhibitor, which has been demonstrated to treat melanoma-associated depression,²¹ as a promising therapeutic option for melanoma, particularly in cases involving BRAF^V600E mutation and resistance to targeted therapy. These findings offer a paradigm for repurposing antidepressants as dual-function therapeutics in cancer care.

It is imperative to highlight that our high-throughput screening of FDA-approved antidepressants reveals that a certain number of them can exert either tumor-promoting or tumor-suppressing effects. This underscores the importance of the careful selection of antidepressants in patients with cancer, as inappropriate choices could exacerbate tumor progression. Given the heterogeneity of the tumors,³⁵ antidepressants should be tested to clarify their effects on the tumor, at least using the patient-derived organoids, before determining the specific antidepressant to be used in oncology patients.

In this study, we revealed that PH served as an unrecognized inducer of pyroptosis, a form of immunogenic cell death characterized by membrane rupture and inflammatory signaling,³⁶ to exert its anti-melanoma effects. Moreover, we demonstrated that pyroptosis induced by PH was mediated through the activation of GSDMB in human melanoma cells and GSDMC in murine models.

Currently, ICIs have become the mainstay of melanoma therapy, but the response rates remain limited.^37,38 In addition, resistance to BRAFi/MEKi in BRAF^V600E-mutant melanoma often results in cross-resistance to immunotherapy, a consequence of the tumor developing an immune-evasive microenvironment.^23,24 Inspired by the fact that inducing pyroptosis triggers a pro-inflammatory TME and thereby boosts immunotherapy response,²⁵ we sought to confirm the ability of pyroptosis induced by PH to potentiate host anti-tumor immunity in DTR melanoma. Our findings revealed that PH-induced pyroptosis enhanced intratumoral CD8⁺ T cell infiltration, which consequently improved the efficacy of the PD-1 blockade, indicating that PH may offer a promising avenue for recurrent melanoma patients to benefit from immunotherapy.

Serotonergic signaling is critical for an array of biological processes ranging from neurotransmission to gastrointestinal motility and hormone release. Prior to this year, research has indicated that intracellular 5-HT has the potential to induce histone serotonylation, consequently leading to ependymoma tumorigenesis.³⁹ Previous studies have indicated that 5-HT plays a crucial role in the proliferation of melanoma cells.⁴⁰ Nevertheless, the precise mechanism remains elusive, and it is unclear whether 5-HT-mediated epigenetic remodeling is involved. Our findings indicate that intracellular 5-HT is essential for maintaining histone serotonylation enrichment and the subsequent transcriptional activation of DNA repair-associated genes. This discovery highlights the previously unrecognized pivotal role of 5-HT in sustaining genomic integrity in melanoma. Once 5-HT reuptake is disturbed, DNA repair capacity is impeded. Consistently, disruption of reuptake of 5-HT blunts the major pathways of DNA damage repair in melanoma cells, which in turn triggers genomic instability. Under these conditions, protein homeostasis is imbalanced in melanoma cells, resulting in substantial protein aggregation and consequent ER stress. This process ultimately leads to the onset of pyroptosis.

In conclusion, our findings illustrate that PH represents a promising approach for melanoma therapy, with the potential for multiple clinical applications. Furthermore, by establishing a link between 5-HT-based epigenetic remodeling and DNA damage repair, we present a comprehensive framework for understanding the role of serotonin in melanoma progression.

Limitations of the study

Duplicated GSDMC transcripts might be functionally compensated for the lack of GSDMB in mice. However, the upstream mediators responsible for GSDMB cleavage in human cells and GSDMC cleavage in murine cells may not be interchangeable. Although we experimentally revealed the underlying mechanism of GSDMC cleavage upon PH treatment and established the causal link between ER stress and GSDMB-dependent pyroptosis upon 5-HT deprivation, the specific molecule that mediates ER stress-triggered GSDMB cleavage under these conditions requires further exploration. Identification of this intrinsic driver of GSDMB activation will fill this gap in the field of pyroptosis. Of note, H3Q5ser antibodies are likely to react with acetylation marks, especially H3K27ac and H3K56ac. Besides proper controls and relevant rescue experiments, high-throughput sequencing, including H3K27ac- and H3K56ac-ChIP-seq, should be performed to confirm that the observed alterations in H3Q5ser signals were not the contaminated signals of histone acetylation. Meanwhile, the clinical implication of 5-HT reuptake activities in BRAF^V600E-mutated melanomas requires further investigation in a larger cohort. Additionally, the in vivo experimental results were primarily obtained using xenograft or allograft models in 4- to 6-week-old female mice. Whether factors such as age and gender constrain the therapeutic efficacy of PH against melanoma requires further clarification.

Resource availability

Lead contact

Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Zhicheng Gong (jichan@jiangnan.edu.cn).

Materials availability

Plasmids generated in this study will be made available on request.

Data and code availability

• •The sequencing data from RNA-seq and ChIP-seq are available from the GEO accession numbers GSE282621 and GSE282619.
• •This paper does not report original code.
• •Any additional information required to reanalyze the data reported in this work paper is available from lead contact upon request.

Acknowledgments

The authors thank Associated Professor Kaisa Cui (Affiliated Hospital of Jiangnan University) for the bioinformatics analysis, Dr. Xuechun Wang (UT Southwestern Medical Center) for proofreading the manuscript, and Associated Professor Fei Chen for drawing the graphical abstract using Biorender.com (agreement no. DC28VE9I58). This work was supported by the National Natural Science Foundation of China (82303115 and 82503908), China Postdoctoral Science Foundation (2024M751162 and 2025MD784141), the Top Talent Support Program for Young and Middle-aged People of Wuxi Health Committee (HB2023060), and Wuxi Science and Technology Innovation and Entrepreneurship Fund “Light of Taihu Lake” Science and Technology Tackling Project (K20231055).

Author contributions

Z.G. conceived and supervised this project. Z.G., X.L., A.L., and L.C. co-designed the experiments. A.L. and X.L. performed most of the experiments. S.X., X.S., J.F., and J.L. supervised some experiments. A.L. analyzed RNA-seq and ChIP-seq data. Z.G., X.L., and A.L. wrote the manuscript with input from all authors.

Declaration of interests

The authors declare no competing interests.

STAR★Methods

Key resources table

REAGENT or RESOURCE	SOURCE	IDENTIFIER
Antibodies
rabbit monoclonal anti-CD8-alpha (Multiplex fluorescence immunohistochemistry)	CST	Cat# 98941; RRID: AB_2756376
rabbit polyclonal anti-Granzyme B (Multiplex fluorescence immunohistochemistry)	ABclonal	Cat# A2557; RRID: AB_2764445
rabbit polyclonal anti-HSP90 (Immunoblotting)	ABclonal	Cat# A5027; RRID: AB_2863419
rabbit polyclonal anti-GRK2 (Immunoblotting)	ABclonal	Cat# A4443; RRID: AB_2863276
rabbit polyclonal anti-Gasdermin B (Immunoblotting)	ABclonal	Cat# A25770
rabbit polyclonal anti-Gasdermin D (Immunoblotting)	ABclonal	Cat# A23755
rabbit polyclonal anti-GSDMC (Immunoblotting)	ABclonal	Cat# A14550; RRID: AB_2769694
rabbit polyclonal anti-GSDMA (Immunoblotting)	ABclonal	Cat# A22624
rabbit polyclonal anti-GSDME (Immunoblotting)	Proteintech	Cat# 13075-1-AP; RRID: AB_2093053
rabbit polyclonal antibody-Histone H3 (Tri-Methyl Lys4, Serotonyl Gln5) (ChIP-Seq)	Beyotime	Cat# AG3850,
rabbit polyclonal anti-PERK (Immunoblotting)	Proteintech	Cat# 24390-1-AP; RRID: AB_2879521
rabbit polyclonal anti-IRE1 (Immunoblotting)	Proteintech	Cat# 27528-1-AP; RRID: AB_2880899
rabbit polyclonal anti-ATF6 (Immunoblotting and Immunofluorescence)	Proteintech	Cat# 24169-1-AP; RRID: AB_2876891
rabbit polyclonal anti-Phospho-IRE1-S724 (Immunoblotting)	ABclonal	Cat# AP1442; RRID:AB_3683584
rabbit polyclonal anti-Phospho-PERK-T982 (Immunoblotting)	ABclonal	Cat# AP1501; RRID: AB_2942102
rabbit monoclonal anti-H3K4me3 (ChIP-Seq)	CST	Cat# 9751S; RRID: AB_11213050
Goat anti-Mouse IgG (H + L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 488 (Immunofluorescence)	Invitrogen	Cat# 1 A-11017; RRID: AB_2534084
Goat anti-Rabbit IgG (H + L) Highly Cross-Adsorbed Secondary Antibody, Alexa Fluor™ 568 (Immunofluorescence)	Invitrogen	Cat# A11036; RRID: AB_10563566
mouse monoclonal anti-γH2AX (Immunoblotting and Immunofluorescence)	millipore	Cat# 05-636-I; RRID: AB_2755003
rabbit polyclonal anti-53BP1 (Immunoblotting)	ABclonal	Cat# A5757; RRID: AB_2766511
In vivo monoclonal anti-IgG2α (clone 2A3)	BioXcell	Cat# BE0089; RRID: AB_1107769
In vivo monoclonal anti-mouse PD-1 (CD279) (clone RMP1-14)	BioXcell	Cat# BE0146; RRID: AB_10949053
rabbit polyclonal anti-XRCC2 (Immunoblotting)	ABclonal	Cat# A1800; RRID: AB_2763839
rabbit monoclonal anti-POLD1 (Immunoblotting)	ABclonal	Cat# A4218; RRID: AB_2863211
rabbit polyclonal anti-Caspase-8 (Immunoblotting)	ABclonal	Cat# A0215; RRID: AB_2757029
rabbit polyclonal anti-Caspase 8/P43/P18 (Immunoblotting)	Proteintech	Cat# 13423-1-AP; RRID: AB_2068463
rabbit recombinant anti-FEN1 (Immunoblotting)	Proteintech	Cat# 84916-5-RR
rabbit polyclonal anti- GSDMC (Immunoblotting)	Proteintech	Cat# 30469-1-AP; RRID: AB_3086329
rabbit monoclonal anti-GSDMB (Immunoblotting)	Abcam	Cat# ab215729; RRID: AB_2909483
rabbit recombinant anti-TPH1 (Immunoblotting)	ABclonal	Cat# A1569; RRID: AB_2763100
rabbit recombinant anti-SLC6A4 (Immunoblotting)	ABclonal	Cat# A23887; RRID: N/A
TruStain FcXTM (anti-mouse CD16/32) (Flow cytometry)	BioLegend	Cat# 101319; RRID: AB_1574973
APC/Cyanine7 anti-mouse CD45 (Flow cytometry)	BioLegend	Cat# 103116; RRID: AB_312981
PE anti-mouse IFN-γ (Flow cytometry)	BioLegend	Cat# 505808; RRID: AB_315401
APC anti-mouse CD45 Recombinant (Flow cytometry)	BioLegend	Cat# 157605; RRID: AB_2876537
PerCP/Cyanine5.5 anti-mouse CD3ℇ (Flow cytometry)	BioLegend	Cat# 100328; RRID: AB_893318
Brilliant Violet 421TM anti-mouse CD4 (Flow cytometry)	BioLegend	Cat# 116023; RRID: AB_2800579
FITC anti-mouse CD8α (Flow cytometry)	BioLegend	Cat# 100706; RRID: AB_312745
PE anti-human/mouse Granzyme B Recombinant (Flow cytometry)	BioLegend	Cat# 372208; RRID: AB_2687032
APC anti-mouse Perforin (Flow cytometry)	BioLegend	Cat# 154404; RRID: AB_2721465
FITC Rat Anti-Mouse CD4 (Flow cytometry)	BD	Cat# 553729; RRID: AB_395013
Bacterial and Virus Strains
E coli Trans109	Lab preserve	N/A
Chemicals, Peptides, and Recombinant Proteins
Paroxetine Hydrochloride	MCE	Cat# HY-B0492
Antidepressant Compound Custom Library	TargetMol	Rack ID:PHD163731/PHD163732/PHD163733
Puromycin	Gibco	Cat# a1113803
Z-DEVD-FMK	MCE	Cat# HY-12466
Serotonin	MCE	Cat# HY-B1473A
Necrosulfonamide	MCE	Cat# HY-100573
Ferrostatin-1	MCE	Cat# HY-100579
protein A/G magnetic beads	MCE	Cat# HY-K0202
16% formaldehyde methanol-free	Thermo	Cat# 28908
Protamine sulfate	AMRESCO	Cat# 0382
Protease Inhibitor Cocktail (PI)	APExBIO	Cat# K1007
Disuccinimidyl glutarate (DSG)	Proteochem	Cat# C1104
Crystal violet	Sigma	Cat# C6158-100G
Cell Counting Kit-8	Targetmol	Cat# C0005
Ampicillin sodium	Sangon	Cat# A610028-0025
N,N,N′,N-Tetramethylethylenediamine (TEMED)	Sigma	Cat# T7024
Penicillin G sodium salt	Sangon	Cat# A600135-0100
Streptomycin sulfate	Sangon	Cat# A610494-0250
FastDigest BshTI	Thermo Scientific	Cat# FD1464
FastDigest EcoRI	Thermo Scientific	Cat# FD0275
30% Acr-Bis (29:1)	Biosharp	Cat# BL513B
Ammonium persulfate	Sigma	Cat# A3678
Luminol	Sigma	Cat# A8511
4-lodophenylboronic	Sigma	Cat# 471933
Dimethyl sulfoxide	MCE	Cat# HY-Y0320
Dulbecco’s Modified Eagle Medium	Gibco	Cat# 12800-017
Fetal Bovine Serum	Gemini	Cat# 900-108
PageRuler Prestained Protein Ladder	Thermo Scientific	Cat# 26617
GoldBand 1 kb DNA ladder	Yeasen	Cat# 10510-A
YeaRed Nucleic Acid Gel Stain	Yeasen	Cat# 10202ES76
NON-Fat Powdered Milk	Sangon	Cat# A600669-0250
Agarose M	Sangon	Cat# A610013-0250
Agar	BioFroxx	Cat# 8211GR500
Albumin, Bovine (BSA)	Amresco	Cat# 0332
10% goat serum	Solarbio	Cat# SL038
Dabrafenib (GSK2118436)	Selleck	Cat# S2807
Trametinib (GSK1120212)	Selleck	Cat# S2673
Critical Commercial Assays
BCA Protein Assay Kit	TIANGEN	Cat# PA115
annexin V-FITC/PI apoptosis detection kit	BD	Cat# 556547
Hematoxylin and Eosin	BOSTER	Cat# AR1180-1
Multiplex Fluorescence Immunohistochemistry Kit-Four-color TSA-RM-275	Panovue	Cat# 10001100020
QIAquick Spin Columns	Qiagen	Cat# 28106
FastPure Plasmid Mini Kit-BOX2	Vazyme	Cat# DC201-01
NucleoBond Xtra Midi	Macherey-Nagel	Cat# 740410.50
Immobilon-P Transfer Membrane	Millipore	Cat# IPVH00010
VAHTS Universal DNA Library Prep Kit for Illumina V3	Vazyme	Cat# ND607
PROTEOSTAT Aggresome Detection Kit	Enzo Life Sciences	Cat# ENZ-51035-K100
ST/5-HT (Serotonin/5-Hydroxytryptamine) ELISA Kit	Elabscience	Cat# E-EL-0033
LDH Cytotoxicity Assay Kit	Beyotime	Cat# C0016
Deposited Data
RNA-seq	This paper	GEO: GSE282621
ChIP-seq	This paper	GEO: GSE282619
Experimental Models: Cell Lines
HEK293T	ATCC	CRL-3216
A2058	Gifted from Han You	CRL-11147
A375	ATCC	CRL-1619
YUMM1.7	ATCC	CRL-3363
UACC903	Gifted from Han You	N/A
LX-2	Gifted from Liang Gong	N/A
GSE-1	Gifted from Liang Gong	N/A
BEAS-2B	Gifted from Liang Gong	N/A
HHL5	BLUEFBIO	BFN6072012687
NCM460	Gifted from Kaisa Cui	N/A
HMC3	ProCell	CL-0620
HT22	Gifted from Ning Bai	N/A
MEF	This paper	N/A
Experimental Models: Organisms/Strains
BALB/c nude mice	TaiYiZhengHe	N/A
C57BL/6J mice	TaiYiZhengHe	N/A
Oligonucleotides
shGRK2#1-Human (5′—3′): AGCGATAAGTTCACACGGTTT	This paper	N/A
shGRK2#2-Human (5′—3′): GAGCGATAAGTTCACACGGTT	This paper	N/A
shGSDMB#1-Human (5′—3′): GCCTTGTTGATGCTGATAGAT	This paper	N/A
shSLC6A4#1-Human (5′—3′): CCCTCTGTTTCTCCTGTTCAT	This paper	N/A
shTPH1#1-Human (5′—3′): CTGTGAATCTACCAGATAATT	This paper	N/A
shGSDMA#1- Human (5′—3′): GGGCTACAGGGATCCATAAAT	This paper	N/A
shGSDMC#1- Human (5′—3′): GCTGTACGATAGCAGTAGTGT	This paper	N/A
shGSDMD#1- Human (5′—3′): CCTTCTCTTCCCGGATAAGAA	This paper	N/A
shGSDME#1- Human (5′—3′): GCATGATGAATGACCTGACTT	This paper	N/A
shGSDMA#1-Mouse(5′—3′): CCAACAGAAGCTTCTAGTAAA	This paper	N/A
shGSDMD#1- Mouse (5′—3′): GATTGATGAGGAGGAATTAAT	This paper	N/A
shGSDME#1- Mouse (5′—3′): CGAGCGTTGTCAATGGGTTAT	This paper	N/A
shGSDMC#1- Mouse (5′—3′): CTCACTTCTGCTAACACTAAG	This paper	N/A
shCaspase8#1- Mouse (5′—3′): GTGAATGGAACCTGGTATATT	This paper	N/A
Primer-QRT-PCR
GAPDH-F:GGAGCGAGATCCCTCCAAAAT	This paper	N/A
GAPDH-R:GGCTGTTGTCATACTTCTCATGG	This paper	N/A
POLD1-F:CAGTGCCAAGGTGGTGTATGG	This paper	N/A
POLD1-R:CTTGCTGATAAGCAGGTATGGG	This paper	N/A
XRCC2-F:TGCTTTATCACCTAACAGCACG	This paper	N/A
XRCC2-R:TGCTCAAGAATTGTAACTAGCCG	This paper	N/A
FEN1-F:CACCTGATGGGCATGTTCTAC	This paper	N/A
FEN1-R:CTCGCCTGACTTGAGCTGT	This paper	N/A
Primer- ChIP-qPCR
FEN1-ChIP-forward (5′—3′): ATTCGCAGTCAAGAGGTCCC	This paper	N/A
FEN1-ChIP-reverse (5′—3′): ACACGACCCCATAAGGTGGA	This paper	N/A
XRCC2-ChIP-forward (5′—3′): CCGGCATCTCAGACGCGTCA	This paper	N/A
XRCC2-ChIP-reverse (5′—3′): TTTACCAAACCGTTCCTCTC	This paper	N/A
POLD1-ChIP-forward (5′—3′): AGGGGTCACGGCGGCGTAGG	This paper	N/A
POLD1-ChIP-reverse (5′—3′): AGTCATGATGGGGAACTAGA	This paper	N/A
Recombinant DNA
pLV-EGFP-TPH1	This paper	N/A
pLV-LIC-Cherry	This paper	N/A
PLKO.1-puro-vector	This paper	N/A
Software and Algorithms
GRAPHPAD prism 9	open source	https://www.graphpad.com/
Flow Jo	open source	https://www.flowjo.com
Modfit	open source	http://www.vsh.com/products/mflt/index.asp
Image Lab	open source	https://www.bio-rad.com/zh-cn/product/image-lab-software
Bowtie2	Langmead and Salzberg41	http://bowtie-bio.sourceforge.net/bowtie2/index.shtml
HOMER	Dr. Christopher K. Glass Lab	http://homer.ucsd.edu/homer/
R software	The R Foundation	https://www.r-project.org/
DESeq2	Love et al.42	https://bioconductor.org/packages/release/bioc/html/DESeq2.html
Java TreeView	Saldanha43	http://jtreeview.sourceforge.net
Fastp	Chen et al.44	https://github.com/OpenGene/fastp
Hisat2	Kim et al.45	http://daehwankimlab.github.io/hisat2/
Samtools	Li et al.46	https://github.com/samtools/
Cufflinks	Trapnell et al.47	http://cole-trapnell-lab.github.io/cufflinks/
Deeptools	Ramirez et al.48	https://deeptools.readthedocs.io/en/develop/
Picard	N/A	http://broadinstitute.github.io/picard/
Intergrative Genomics Viewer (IGV)	Broad Institute	http://software.broadinstitute.org/software/igv/
LAS X	Leica	https://www.leica-microsystems.com/products/microscope-software/p/leica-las-x-ls/
ImageJ	ImageJ	https://imagej.nih.gov/ij/
CAPS version 1.0.1	open source	https://www.bio-launching.com/s02/fileinfo/2013/06/28/809751.html

Experimental model and study participant details

Animal models

All mouse experiments were approved and followed the guidelines established by the Institutional Animal Care and Use Committee of Jiangnan University (JN.NO20240830b0401110(392)). 4–6-week-old female BALB/c nude mice and C57BL/6J mice were obtained from TaiYiZhengHe Biological company (China). All mice were maintained under specific pathogen-free conditions with a 12 h light/12 h dark cycle, at 22 ± 2°C with humidity of 50% ± 5% and fed with a standard mouse chow diet at the Jiangnan University Laboratory Animal Center. Mouse experiments were performed in strict accordance with proper animal practices as defined by the Jiangnan University Laboratory Animal Center.

Cell lines

STR profiling was used to authenticate the cells employed in this study, and mycoplasma contamination testing yielded negative results. The A375, YUMM1.7, HEK293T cell lines were obtained from the ATCC. The UACC903 and A2058 cell lines were kindly provided by Professor Han You (School of Life Sciences, Xiamen University). MEF cells were isolated from E12.5-E18.5 embryos. Briefly, embryos were decapitated, eviscerated, and limbs/tail removed. The remaining tissue was minced, digested in trypsin (20 min), dissociated, and centrifuged. After supernatant removal, cells were resuspended in medium and plated in 10 cm plate. HT22 was a kind gift from Ning Bai (Health Sciences Institute, China Medical University). LX-2, GSE-1 and BEAS-2B cells were gifted from Liang Gong (Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences). HHL5 was obtained from BLUEFBIO (BFN6072012687). NCM460 cells were kindly provided by Kaisa Cui (Wuxi Medical School, Jiangnan University). HMC3 cells were purchased from ProCell (CL-0620). GSE-1 and NCM460 cells were cultured in RPM1640 (Gibco, 61870036) supplemented with 10% fetal bovine serum (FBS Gemini, 900-108). HMC3 cells were maintained in MEM containing NEAA (Procell, PM150410) supplemented with 10% fetal bovine serum. Other cells were maintained in culture in DMEM (Gibco, 12800-017), supplemented with 10% FBS, at 37°C with a saturated humidity atmosphere comprising 95% air and 5% CO₂. To generate DTR cell lines, parental cells were treated with 250 nM dabrafenib (Selleck Chemicals, S2807) and 10 nM trametinib (Selleck Chemicals, S2673) for approximately 3 months. The medium containing DT was refreshed every 3 days. DTR cells were cultured with DT.

Method details

RNA interference

Gene silencing was conducted using the lentiviral vector pLKO.1-puro, and the shRNA sequences for the target genes are listed in the key resources table.

Drug screening analysis

To identify small molecules that could affect melanoma cell viability, 90 FDA-approved antidepressant compounds were purchased from TargetMol for a drug screening. In detail, A375 cells were seeded at a density of 2000 cells per well in 96-well plates and cultured for 24 h, then treated with each compound (10 μM) for 48 h. Cell viability was measured using the Cell Count Kit-8 (TargetMol, C0005) according to the manufacturer’s instructions.

Clonogenic assay and CCK8 assays

6,000 to 10,000 cells were seeded in 6-well plates and cultured for 6 to 12 days. The recommended working concentration for PH was 0.5–7.5 μM, and 5-HT (MCE, HY-B1473A) was used in doses ranging from 15 to 25 μM. After washing with Phosphate-Buffered Saline (PBS), the cells were fixed with formaldehyde (SCR, 10014118), stained with 2% crystal violet (Sigma-Aldrich, C6158-100G) and imaged with a digital scanner. Crystal violet was dissolved in a destaining solution (50% methanol and 10% acetic acid) and measured at OD₅₆₀. Cell viability was measured using the Cell Count Kit-8 (Targetmol, C0005) according to the manufacturer’s instructions. The recommended working concentration for PH was 1–12.5 μM and the doses of 5-HT (MCE, HY-B1473A) are ranging from 10 to 30 μM.

Animal studies

All mice were kept under specific pathogen-free conditions with a 12-h light/12-h dark cycle, at 22 ± 2°C with humidity of 50% ± 5%, and fed a standard mouse chow diet at the Jiangnan University Laboratory Animal Center. 4–6-week-old female BALB/c nude mice and C57BL/6J mice were obtained from TaiYiZhengHe Biological company. A total of 2 × 10⁶ A375 cells or 2 × 10⁵ YUMM1.7^DTR cells were injected subcutaneously into the posterior flanks of nude mice or C57BL/6J mice. Tumor volume was monitored at two-to three-day intervals using calipers and calculated according to the following formula: (width² × length)/2 (mm³). For PH, MCE, HY-B0492 treatment, once tumor volumes had reached an approximate range of 50–200 mm³, mice were randomly assigned to treatment groups and administered the drug. PH was dissolved in Phosphate-Buffered Saline (PBS) and the doses and treatment frequency were tested based on the clinical use. Finally, PH was used in vivo at a concentration of 25 mg/kg body weight, and administered intraperitoneally (i.p.) once daily. The antibodies IgG2a (BioXcell, clone 2A3, BE0089) and PD-1 (BioXcell, clone RMP1-14, BE0146) were diluted in PBS and administered via intraperitoneal injection at a dose of 200 μg every 3 days.

Co-culture assay

First, cDNA encoding TPH1 was amplified and cloned into the pLV-EGFP vector. Next, we generated a stable A375 cell line overexpressing GFP-TPH1 via lentiviral transduction of the indicated gene. Concurrently, Cherry-labeled A375 cells were generated using lentivirus encoding Cherry. The two cell populations were then mixed, and 200,000 cells were seeded into 6-well plates for biological replicates. After treating the mixed cells with vehicle or PH for 48 h, the number of Cherry-labeled cells was quantified using flow cytometry (BD).

Immunoblotting (IB)

Cells were lysed with RIPA lysis buffer (50 mM Tris pH 8.0, 150 mM NaCl, 1% Triton, 0.1% SDS, 1 mM EDTA, 1% sodium deoxycholic acid, 1 mM Na3VO4, 10 mM NaF) containing protease inhibitor and phenylmethylsulfonyl fluoride (PMSF). The Protein concentration of each sample was determined by using the BCA kit (TIANGEN Biotech Co., Ltd.) according to the manufacturer’s instructions. Equivalent quantities of Cell lysates were fractionated using SDS-PAGE and then electro transferred onto polyvinylidene fluoride (PVDF) membranes (Millipore). Subsequently, the membranes were probed with the specific antibodies. The membranes were first incubated with various primary antibodies, followed by treatment with secondary antibodies linked to horseradish peroxidase. Quantitative densitometry analysis was performed with image analysis software (Quantity One, BioRad). The Antibodies used are listed in the key resources table.

Flow cytometry

Apoptosis assay: After indicated treatments, cells were collected and washed with PBS twice, then stained with an annexin V-FITC/PI apoptosis detection kit (BD). The stained cells were analyzed by a Quanteon/ACEA flow cytometer along with FlowJo software (Version 10.5.3, TreeStar).

To characterize tumor-infiltrating T cells, we prepared single-cell suspensions from fresh tumor specimens using enzymatic digestion. Tumor tissues were dissociated with an enzyme cocktail containing 1 mg/mL fetal collagenase D and 0.15 mg/mL DNase at 37°C for 1 h. The resulting cell suspension was then subjected to Percoll density gradient centrifugation (30–70% discontinuous gradient) to enrich for immune cell populations. For immunophenotyping, cells were incubated with fluorochrome-conjugated antibodies targeting surface markers at 4°C for 30 min, followed by washing steps to remove unbound antibodies.

For intracellular staining of GZMB, PRF1 and IFN-γ, cells were activated with Leukocyte Activation Cocktail (BD Biosciences, #550583), for 5 h at 37°C in 5% CO₂. Following stimulation, cells were Fixed with Fixation Buffer (BioLegend, #420801) for 20 min at room temperature. Permeabilized using Intracellular Staining Permeabilization Wash Buffer (BioLegend, #421002). After surface staining, samples were then subjected to intracellular staining using indicated flow antibodies at 4°C for 60 min.

To analyze peripheral T cells, single-cell suspensions were prepared from fresh spleen tissues through mechanical dissociation using 70 μm cell strainers or from peripheral blood via density gradient centrifugation (400 g, 20 min). Erythrocytes were subsequently lysed with erythrocyte lysate (Beyotime, #C3702) for 5 min at room temperature, followed by two washes with PBS containing 2% FBS. Surface marker staining and subsequent flow cytometry analysis were performed using the same protocol established for intratumoral T cell characterization. Samples were acquired on FACSymphony A1 flow cytometers, with data analysis conducted using FlowJo software (v10.5.3; TreeStar), applying compensation matrices generated from single-stained controls and standardized gating strategies for T cell subset identification.

Protein aggregation detection assay

The PROTEOSTAT Aggresome Detection Kit (Enzo Life Sciences, ENZ-51035-K100) was employed to identify aggregated proteins in the cellular samples. Aggresome detection was conducted according to the manufacturer’s instructions. Briefly, the A375 cells cultured on glass slides were first rinsed with PBS, followed by fixation in 4% formaldehyde for half an hour at ambient temperature. Subsequently, the cells were permeabilized using a Permeabilizing buffer (0.5% Triton X-100 and 3 mM EDTA) for 30 min on ice, with a gentle shaking motion. Finally, the cells were incubated with the PROTEOSTAT dye at a 1:20,000 dilution for 30 min at RT. The nuclei were counterstained with DAPI. The fluorescence images were acquired using a laser scanning confocal microscope (Zeiss LSM 980+Airyscan2) and quantified using the ImageJ software.

Immunofluorescence

Cells on glass coverslips were fixed in 4% formaldehyde for 15 min at RT, permeabilized for 5 min with 0.2% (vol/vol) Triton X-100 in PBS, washed with PBS, and incubated with primary antibodies for 60 min at RT or overnight at 4°C. Then, the coverslips were incubated with secondary antibodies for 60 min at RT, washed with PBS, and mounted using Vectashield mounting reagent with DAPI (Vector Laboratories). The fluorescence images were acquired using a laser scanning confocal microscopy (Zeiss LSM 980+Airyscan2) and quantified using the ImageJ software.

Immunohistochemistry (IHC)

The deparaffinization process of formalin-fixed paraffin-embedded tissue sections (4 μm thickness) was initiated with dimethylbenzene treatment followed by rehydration through a graded alcohol series. Microwave-Assisted Antigen Retrievall was subsequently performed using either sodium citrate buffer.

For multiplex fluorescent immunohistochemical staining, tissue sections were washed three times with 1 × TBST (Tris-buffered saline with 0.1% Tween 20) and subsequently blocked with a blocking buffer containing 3% bovine serum albumin (BSA), 0.3% Triton X-100, and 10% goat serum (Solarbio, #SL038) for 1 h at room temperature. Primary antibodies diluted in blocking buffer were then applied and incubated overnight at 4°C. After three additional TBST washes, sections were incubated with fluorescent-conjugated secondary antibodies (Panovue, #10015001006) for 30 min at room temperature, followed by tyramide signal amplification using the TSA-RM-275 kit (Panovue, #0001100020) according to the manufacturer’s protocol. High-resolution whole-slide imaging was performed using a Nikon Ti-E ds-Ri2 NY USA digital slide scanning system with appropriate fluorescence filter sets for each channel.

Hematoxylin and eosin (HE) staining

Formalin-fixed, paraffin-embedded sections (4 μm) of major murine organs were subjected to standard HE staining. Briefly, the sections were deparaffinized in xylene and rehydrated through a graded ethanol series (100%–70%), followed by staining with Mayer’s hematoxylin (1 min) and eosin Y (1 min), with intervening washes with distilled water. After dehydration in ethanol and clearing in xylene, the slides were mounted with neutral resin and imaged under a bright-field microscope (Nikon).

Comet assay

Cells were harvested as a single-cell suspension, then combined with 0.5% low-melting-point agarose (Gibco) into PBS. This mixture was subsequently cast onto a microscope slide that had been pre-layered with a 1% normal-melting-point agarose (Invitrogen) coating. The cells were incubated with lysis buffer (2.5 M NaCl, 100 mM EDTA, 10 mM Tris, and 1% Triton X-100) overnight and then thoroughly rinsed with a neutralizing buffer (0.4 M Tris–HCl at a pH of 7.4). Electrophoresis was performed in an alkaline electrophoresis solution at 25 V for 25 min and fixed in 96% ethanol. DNA was stained using SYBR Green I (Molecular Probes) and visualized using fluorescence microscope (Olympus CX33). The comet tail moment (tail moment = tail DNA% × tail length) was measured by CAPS (comet assay software project) version 1.0.1 software to evaluate the degree of DNA damage in cells.

5-Hydroxytryptamine (5-HT) levels assay

A375 and YUMM1.7^DTR cells were treated with different concentrations of PH for 24 h. Cells were trypsin-digested, counted, resuspended in PBS at a final concentration of 5 × 10⁶ cells/mL, and then subjected to sonication. Samples from control, standard and experimental groups were analyzed by taking 50 μL of each sample, and the experiments were performed according to the instructions of ST/5-HT (Serotonin/5-Hydroxytryptamine) ELISA Kit (Elabscience, E-EL-0033).

LDH release assay

The activity of LDH released into the cell culture supernatants was detected using an LDH Cytotoxicity Assay Kit (Beyotime, C0016) according to the manufacturer’s protocol.

Chromatin immunoprecipitation (ChIP)

For H3K4me3 (CST, C42D8, 9751S#) and H3-5HT (Beyotime, AG3850) ChIP, cells were fixed with 1% formaldehyde (Thermo, 28908) for 15 min at RT. Cells were then washed with PBS and terminated by the addition of 0.125 M glycine. The pellet was resuspended in 1% SDS lysis buffer after removal of cytosolic proteins with cytosolic buffer (50 mM Tris-HCl pH = 8.0, 0.5% NP-40 and 50 mM NaCl). Chromatin DNA was sheared by sonication to an average size of 300–500 bp. The lysates were incubated with IgG or specific antibodies overnight at 4°C and then incubated with protein A/G magnetic beads (MCE, HY-K0202) for 4 h. After washing, the eluted protein-DNA complex was reversed by heating at 65°C overnight. The immunoprecipitated DNA was purified on QIAquick Spin Columns (Qiagen, 28106) and high-throughput sequencing was performed. Relative ChIP enrichment was confirmed using qPCR. The primers used for ChIP-qPCR are listed in the key resources table.

Quantitative RT-PCR

Total RNA was isolated from the samples using TRIzol reagent, following the manufacturer’s instructions. cDNA was prepared using RevertAid Reverse Transcriptase. Quantitative PCR was performed using the StepOnePlus real-time PCR system (Applied Biosystems Inc., Foster City, CA, USA), which measures real-time SYBR Green fluorescence, and then calculated using the comparative Ct method, with the expression of GAPDH as an internal control. The primers used for qRT-PCR are listed in the key resources table.

Bioinformatic analyses

Gene expression profiles of patients with the BRAF^V600E mutation (n = 48) were obtained from the TCGA database, and gene ontology analysis was performed using Metascape (http://metascape.org/gp/index.html). The expression pattern of TPH1 in A375 was analyzed using GSE162362 dataset.⁴⁹ scRNA-seq analysis was based on a dataset with annotated clusters (GSE229908).⁵⁰ As previously described,⁵¹ the density of TPH1⁺/SLC6A4⁺ cells and pyroptosis signature⁵² were visualized using the Nebulosa package in R software (https://doi.org/10.18129/B9.bioc.Nebulosa).

ESTIMATE tumor purity and immune score analyses were conducted as previously described⁵³ using R software with a gene expression matrix in patients with BRAF^V600E mutation from the TCGA database.

Quantification and statistical analysis

All in vitro data are presented as mean ± SD of three independent experiments with similar results unless otherwise indicated. All in vivo data are presented as mean ± SEM, as described in corresponding legends. Quantification of immunostaining intensities was achieved using ImageJ on image acquired as above mentioned. Statistical significance was determined by unpaired Student’s t test using Prism 9.3.1 software (GraphPad Software). p-values and sample sizes are presented in main and supplementary figure legends. p < 0.05 was considered significant.

Published: January 5, 2026